Participation in psychoeducation groups for families with a member with severe mental illness has consistently been shown to be robustly associated in rigorously conducted efficacy studies with a variety of positive gains for patients and their families. Recently, McFarlane (1993, 1995) has successfully disseminated his well-established Multifamily Psychoeducation model from inpatient settings to other clinical mental health settings. Development and empirical validation of a community model that preserves the benefits derived from clinical efficacy trials is therefore the important next step to making these programs available to families in communities. Therefore, the purpose of the proposed project is to test our Family Responsive Psychoeducation Group (FRPG) model, an innovation adapted from McFarlane's model, in a real-world, non-clinical community setting (the Alliance for the Mentally Ill [NAMI]). Our adaptation significantly advances the existing state-of-the-art with incorporation of new elements, in that FRPG: (a) is not limited just to families with only one particular type of mental illness in its ill member; (b) provides high fidelity of the intervention through use of a standardized manual, while retaining a flexible curriculum that is responsive to a wide range of family group concerns; and (c) builds upon previously developed methods aimed at recruiting families through one day workshops (Anderson et al., 1986) and retaining families who are ready and motivated for change. The effectiveness of FRPG will be tested in a community setting (NAMI) through a random assignment experiment comparing individual-and family- level outcomes for families with a member with severe mental illness. Families will be randomly assigned to either (1) FRPG or (2) a "usual care" condition in which individuals access standard NAMI services including their local NAMI family education group. Participants in this study will be recruited from a series of one-day community workshops for families and their ill members aimed at providing basic education about serious mental illness. Data will be collected through: (a) systematic interviews of the ill family member in both experimental conditions at baseline, 12, and 24 months to obtain data about relapse, psychiatric symptom levels, medication compliance, social functioning/quality of life, service, use, and costs; (b) systematic interviews of a key family member (one per family) in both experimental conditions at baseline, 12, and 24 months to obtain data about family well-being, social support, family burden and coping, service use, and family-level costs; (c) surveys of all workshop participants for phase-of-change process and satisfaction with the workshop; and (d) surveys of all FRPG group participants every 2 months at group meetings for satisfaction with the multifamily groups.